An increase in the mammographic density of breast tissue is correlated to a four to six-fold increased risk of developing breast carcinoma, making it the single biggest risk factor for breast carcinoma. Despite this, the molecular mechanism by which breast density links to carcinoma formation is unknown. We have recently shown that in a mouse model of increased collagen deposition, mammary tumor incidence, invasion, and metastasis increase three-fold, suggesting that an increase in collagen per se is part of the mechanism by which increased breast density is correlated with increased risk in humans. The purpose of this proposal is to understand how physical properties of the ECM regulate breast cell behavior. Our hypothesis is that locally dense ECM enhances the formation of matrix adhesions that result in activation of signaling pathways linked to FAK, and that FAK is a central mediator by which matrix density regulates gene expression to promote proliferation and invasion. This hypothesis will be tested in the following Specific Aims: Aim 1: Define changes in the expression of proliferation and metastasis-associated genes regulated by collagen density and alignment. Gene expression of the proliferation and the metastasis gene signatures, will be determined in human breast carcinoma samples, using tissue arrays linked to patient outcome, and with DCIS and microinvasive DCIS. Metastasis genes will be screened by siRNA and shRNA knock-down for their role in mediating invasion into 3D matrices in vitro and in vivo. Aim 2: Test the hypothesis that FAK regulates proliferation and invasion in response to collagen density and alignment Tumors that are FAK-/- do not invade or metastasize in vivo, even in the context of a FAK+/+ stroma. We will use in vitro 3D invasion assays and our in vivo FAK-/- mice to investigate the role of FAK in cell proliferation, matrix alignment and 3D invasion. Conditional FAK knockout mice will be used to determine the role of FAK in enhanced tumor progression in the collagenase-resistant transgenic mouse model (Col1a1tm1Jae) having a dense collagen matrix. Phosphorylation of FAK at pY397 will be determined in mammary tissues from Col1a1 mice, mouse tumor, and human pathologic samples to determine if pY397FAK correlates to dense breast tissue in vivo. Aim 3: Test the link between collagen density, FAK and downstream signaling events, ERK, Src, and PI3K, in regulating cell proliferation and invasion. FAK-/- cells lose activation of ERK and PI3K pathways. Using pharmacologic inhibitors and siRNA/shRNA approaches, we will inhibit each of these pathways and determine their role in mediating the proliferation and invasion of human breast cells into dense and aligned collagen matrices in vitro. Xenografts of human carcinoma, and mice bearing tumors in wt and dense collagen stroma (Col1a1 mouse model) will be treated in vivo with pharmacologic inhibitors to test the role of these molecules in tumor progression in vivo.